Peanut (Arachis hypogaea L.) is a highly nutritious legume that provides energy-dense food. One obstacle for universal inclusion of peanut products in the diet is the prevalence of peanut food allergy. Although approximately 2% of individuals may become sensitized to peanut and experience an allergic reaction upon ingestion (Gupta et al., 2011), a minor fraction of these reactions may be life-threatening. The 2S albumins (Ara h 2, Ara h 6 and Ara h 7) are the least abundant seed storage proteins but Ara h 2 and 6 are the most potent allergens (Zhuang and Dreskin, 2013). Multiple isoforms of each 2S albumin exist and tightly linked gene copies encode each protein. One copy of each gene spans 83 kb of chromosome Arahy.08 and one extra copy of Ara h 6 is present within a 163-kb segment on Arahy.18 of Arachis hypogaea (Table S1). While previously shown that reduced expression of the 2S albumins, achieved through RNAi knockdown, did not affect seed growth and development, low amounts of protein that could possibly incite an allergic reaction were still detected (Chu et al., 2008).

Protein knockouts through induced mutagenesis or discovery of a natural null mutant is an effective way to ensure protein elimination. A null mutant of the β-conglycinin α-subunit gene in soybean was used to backcross breed lines for reduced allergen and improved quality tofu (Song et al., 2014) while a null mutant of the soybean P34 allergenic protein was used in breeding of an improved and low allergen variety (Jeong et al., 2013). CRISPR-Cas gene editing is recognized as an efficient method for creating targeted mutations in multiple alleles and/or genes (Van Eck, 2020). Given the tight linkage of the 2S albumin genes, a multiplex gene editing strategy using CRISPR-Cas9 and two conserved guide RNAs (Figure 1a; Figure S1, Table S2) for each Ara h gene was attempted. We report the successful multiplexed editing of 2S albumin allergen genes of peanut, transmission through the next generation, and protein elimination.

Fifty-two out of 113 total Cas9-positive T₀ plants recovered using methods described in Supporting Information produced pods of varying seed quality and number. T₀ plants were screened for large CRISPR deletions by agarose gel separation of PCR products (Figure 1b; Table S3). Fourteen T₀ plants displayed an Ara h 2 deletion with plants in lines 127 and 214 producing seed. Five T₀ plants showed Ara h 6 deletions with line 45 producing seed. Twelve T₀ plants showed Ara h 7 deletions with line 127 producing seed. Next-generation sequencing of T₁ seed/seedlings and Khufu analysis was used to identify smaller CRISPR-Cas9 edits. To maximize the number of T₁ individuals screened, each sequencing well combined three individuals and amplicons from all target genes. A large variation in the number of occurrences and changes were identified for the seven Ara h genes. Changes for all Arah6 and Arah2_B were limited while numerous edits for Arah7_A and _B and Ara2_A were identified (Table S4). While focusing on edits which created nearby stop codons, many identified edits were not knockouts as they created in-frame deletions of 3, 6 or 9 nucleotides or nonsense mutations which did not create a nearby stop codon. Sanger sequencing confirmed Ara h CRISPR edits for six T₁ individuals as shown in Figure 1, Figure S2 and Tables S5, S6. Individuals without clones for an Ara h gene (nd) may have completely lost the gene which would not be obvious in PCR amplification due to the A and B homoeologs. Clones from line 127, which had both Ara h 2 and Ara h 7 deletions, displayed larger than expected bands which were also cloned and sequenced. These clones showed a 463-bp insertion in Arah2_A and a 206 bp insertion for Arah7_A. The 206 bp Arah7_A insertion was an exact match for the region deleted between the two guides from Arah2_B while the 463 bp Arah2_A insertion had similarity to multiple regions of the Tifrunner. Arahy chromosomes. These complex rearrangements can sometimes occur upon repair of double-strand breaks.

Sanger sequenced T₁ lines were compared against T₀ parents and unsequenced siblings using PCR amplification of the Ara h genes and inheritance of the transgenes using hygromycin resistance gene primers (Figure 1b). Based on amplicon sizes, T₁ offspring looked similar to their T₀ parents, although larger amplicons were generated for multiple offspring. Offspring derived from lines 214 and 45 seem to have lost the transgenes based on lack of hygromycin resistance gene amplification (Figure 1).

Western blots probed with antibodies for Ara h 2 and Ara h 6 confirmed the disappearance of some protein bands (Figure 1c). Protein isolated from line 42 seed lost Arah2_B and retained a wild-type copy of Arah2_A. Protein from multiple sibs of lines 46 and 89 seed and predicted by Sanger sequencing to show knockout of both copies of Ara h 2 was confirmed. Knockouts of Ara h 2 were also found among T₂ seed of lines 127, 45 and 214. The protein analysis of Ara h 6 showed offspring of lines 45 and 46 with a reduction, but not a complete absence, of Ara h 6. No antibodies are available for analysis of Ara h 7.

The 2S albumin genes in peanut, comprised of seven family members for Ara h 2, Ara h 6 and Ara h 7, were successfully targeted for multiplexed gene editing with a polycistronic tRNA-gRNA (Wolabu et al., 2020) to boost the recovery of mutants. Guides were designed to conserved regions of both A- and B-genome copies of each gene, and although the ability of all guides to cleave the target sequences was tested with an in vitro screen, in vivo efficiency varied: Ara h 7>Ara h 2>Ara h 6. Lines 45 and 46 are particularly promising for elimination of most 2S albumin according to the Sanger sequencing data although the western blot still shows a faint band suggesting some possible residual. More accurate and sensitive methods for quantifying 2S albumins, such as the reverse-phase LC–MS/MS spectral counting used in a previous study (Stevenson et al., 2009), would need to be applied to resolve this question. If elimination is confirmed, a test for reduced allergenicity logically would be conducted with a mouse model or an ex vivo basophil degranulation assay perhaps followed by skin prick assay (Ozias-Akins et al., 2009). Finally, given the identification of unexpected PCR products, whole-genome sequencing of promising lines where the transgene has been genetically removed through segregation would be needed to confirm the changes and identify possible gene rearrangements or off-target genome edits by Cas9.

花生（ArachishypogaeaL .）是一种营养丰富的豆类，提供能量密集的食物。普遍将花生产品纳入饮食的障碍之一是花生食物过敏的流行。尽管大约 2% 的人可能对花生过敏并在摄入后出现过敏反应（Gupta等，  2011），但这些反应中的一小部分可能会危及生命。 2S 白蛋白（Ara h 2、Ara h 6 和 Ara h 7）是最不丰富的种子储存蛋白，但 Ara h 2 和 6 是最有效的过敏原（Zhuang 和 Dreskin，  2013）。每个 2S 白蛋白存在多种亚型，并且紧密相连的基因拷贝编码每个蛋白质。每个基因的一个拷贝跨越 83 kb 的染色体 Arahy.08，并且 Ara h 6 的一个额外拷贝存在于花生Arahy.18 的 163 kb 片段中（表 S1）。虽然之前表明通过 RNAi 敲低实现的 2S 白蛋白表达减少不会影响种子的生长和发育，但仍然检测到可能引发过敏反应的少量蛋白质（Chu等人，  2008）。

通过诱导诱变或发现天然无效突变体来敲除蛋白质是确保消除蛋白质的有效方法。大豆β-伴大豆球蛋白α亚基基因的无效突变体用于回交品种系，以减少过敏原并提高豆腐质量（Song等，  2014），而大豆P34过敏蛋白的无效突变体则用于育种改良且低过敏原的品种（Jeong等，  2013）。 CRISPR-Cas 基因编辑被认为是在多个等位基因和/或基因中创建靶向突变的有效方法（Van Eck，  2020）。鉴于 2S 白蛋白基因的紧密连锁，尝试对每个Ara h基因使用 CRISPR-Cas9 和两个保守向导 RNA（图 1a；图 S1，表 S2）的多重基因编辑策略。我们报告了花生 2S 白蛋白过敏原基因的成功多重编辑、下一代传递和蛋白质消除。

_{使用支持信息中描述的方法回收的 113 株 Cas9 阳性 T 0}植物中，有 52 株产生了不同种子质量和数量的豆荚。通过 PCR 产物的琼脂糖凝胶分离来筛选T ₀植物的大 CRISPR 缺失（图 1b；表 S3）。十四个T ₀植物显示出Ara h 2缺失，其中品系127和214中的植物产生种子。五株T ₀植物显示出Ara h 6缺失，其中品系45产生种子。 12 个 T ₀植物显示Ara h 7缺失，其中品系 127 产生种子。 T ₁种子/幼苗的新一代测序和 Khufu 分析用于识别较小的 CRISPR-Cas9 编辑。为了最大化筛选的 T ₁个体的数量，每个测序孔结合了三个个体和来自所有目标基因的扩增子。七个Ara h基因的出现和变化数量存在很大差异。所有 Arah6 和 Arah2_B 的更改都很有限，而 Arah7_A 和 _B 以及 Ara2_A 的大量编辑已被识别（表 S4）。虽然重点关注创建附近终止密码子的编辑，但许多已识别的编辑并不是敲除，因为它们创建了 3、6 或 9 个核苷酸的框内删除或不创建附近终止密码子的无义突变。 Sanger 测序证实了6 个 T _1个体的Arah CRISPR 编辑，如图 1、图 S2 和表 S5、S6 所示。没有Ara h基因 (nd) 克隆的个体可能完全丢失了该基因，由于 A 和 B 同源物，该基因在 PCR 扩增中并不明显。来自品系 127 的克隆具有Ara h 2和Ara h 7缺失，显示出比预期更大的条带，这些条带也被克隆和测序。这些克隆在 Arah2_A 中显示出 463 bp 的插入，在 Arah7_A 中显示出 206 bp 的插入。 206 bp Arah7_A 插入与 Arah2_B 的两个引导之间删除的区域完全匹配，而 463 bp Arah2_A 插入与 Tifrunner 的多个区域相似。阿拉希染色体。这些复杂的重排有时会在双链断裂修复时发生。

使用Ara h基因的 PCR 扩增和使用潮霉素抗性基因引物的转基因遗传，将Sanger 测序的 T _{1系与 T 0}父母和未测序的兄弟姐妹进行比较（图 1b）。根据扩增子大小，T ₁后代看起来与其 T ₀父母相似，尽管为多个后代生成了更大的扩增子。由于缺乏潮霉素抗性基因扩增，源自 214 和 45 系的后代似乎已经失去了转基因（图 1）。

用 Ara h 2 和 Ara h 6 抗体检测的蛋白质印迹证实了一些蛋白质条带的消失（图 1c）。从品系 42 种子中分离的蛋白质丢失了 Arah2_B，并保留了 Arah2_A 的野生型副本。证实来自品系 46 和 89 种子的多个同胞的蛋白质，通过桑格测序预测显示 Ara h 2 两个拷贝均被敲除。在品系 127、45 和 214 的 T ₂种子中也发现了 Ara h 2 的敲除。Ara h 6 的蛋白质分析显示品系 45 和 46 的后代 Ara h 6 减少，但并非完全缺失。抗体可用于分析 Ara h 7。

花生中的 2S 白蛋白基因由Ara h 2、Ara h 6和Ara h 7的七个家族成员组成，成功地利用多顺反子 tRNA-gRNA 进行多重基因编辑（Wolabu等人，  2020）以促进恢复突变体。指导被设计为每个基因的 A 和 B 基因组拷贝的保守区域，尽管所有指导切割靶序列的能力都通过体外筛选进行了测试，但体内效率各不相同：Ara h 7>Ara h 2>阿拉h 6。根据 Sanger 测序数据，第 45 和 46 系特别有希望消除大多数 2S 白蛋白，尽管蛋白质印迹仍然显示微弱的条带，表明可能有一些残留。需要应用更准确和更灵敏的 2S 白蛋白定量方法来解决这个问题，例如先前研究中使用的反相 LC-MS/MS 光谱计数（Stevenson等人，  2009 ）。如果确认消除，则逻辑上将用小鼠模型或离体嗜碱性粒细胞脱颗粒测定进行降低过敏性的测试，可能随后进行皮肤点刺测定（Ozias-Akins等人，  2009）。最后，考虑到意外 PCR 产物的鉴定，需要对转基因已通过分离从基因上去除的有希望的品系进行全基因组测序，以确认变化并鉴定可能的基因重排或 Cas9 的脱靶基因组编辑。